VOLUME 74, NUMBER 10
PHOTOGRAMMETRIC ENGINEERING & REMOTE
SENSING
JOURNAL OF THE AMERICAN SOCIETY FOR PHOTOGRAMMETRY
AND REMOTE SENSING
This month’s cover is provided by Dr. Giorgos Mountrakis, Director of the
Intelligent Geocomputing Lab at the State University of
New York College of Environmental Science and Forestry
(www.esf.edu). The imagery complements this Special Issue
on “Artificial Intelligence in Remote Sensing”. Artificial
intelligence methods often borrow elements from art and
science. The background image contains a 2003 natural
color digital orthoimagery of Syracuse, New York with a
pixel size of 1.0 ft. GSD. The orthoimagery is processed
using various artistic spatial filters.
The center image is produced using an expert-based
system on an April, 2000 Landsat scene from Las Vegas,
NV. The system intelligently balances its complexity to
adjust to the intricacies of the underlying classification.
It selectively and progressively moves from simple classifiers to mathematically complex ones such as neural
networks and decision trees. The image shows the spatial
footprint of each algorithmic approach. The integration of
numerous competing classifi ers offers advanced classification
capabilities; the resulting binary classifi cation of urban
areas is depicted on the top right image. For more information
contact Giorgos Mountrakis at gmountrakis@esf.edu or visit www.aboutgis.com.
A back-propagation multilayer perceptron, self-organizing
map, fuzzy ARTMAP, and gini and entropy univariate decision
trees compared in terms of their ability to cope with small,
unrepresentative, and variable training sets.